About the job
The Industrial Compute team at OpenAI focuses on building and expanding large-scale compute capabilities. This includes work across proprietary data centers, strategic partnerships, and industrial infrastructure. The team’s mission centers on transforming resources, such as power, land, hardware, and operational efficiency, into reliable compute capacity that fuels advanced AI training and inference.
Operating at the crossroads of infrastructure delivery, hardware systems, utilities, supply chain management, and capacity strategy, the team aims to help OpenAI scale its computing resources faster than traditional approaches.
Role overview
The Tokens-as-a-Service (TaaS) Lead will be based in San Francisco and will guide the transformation of industrial-scale infrastructure investments into operational token capacity for OpenAI’s workloads. This role manages complex compute programs, turning raw infrastructure into functional GPU throughput. Collaboration is key: the TaaS Lead partners with teams handling data center delivery, power, networking, hardware deployment, workload enablement, finance, and external organizations to move capacity into productive use quickly.
This position fits someone who excels at linking physical infrastructure delivery to real compute utilization. Success in this role depends on strong systems thinking, demonstrated program leadership, and the ability to hold both internal teams and external partners accountable for outcomes.
What you will do
- Lead Tokens-as-a-Service projects in industrial compute settings, spanning both OpenAI-owned and partner capacities.
- Convert delivered power, space, and hardware into production-ready token throughput.
- Create execution plans covering construction, power-up, rack installation, networking, cluster readiness, and workload onboarding.
- Work closely with infrastructure engineering, hardware, networking, finance, supply chain, and operations teams.
- Direct external providers, EPCs, OEMs, utilities, and strategic partners to deliver on tight timelines.
- Identify and resolve bottlenecks between physical delivery and usable compute output.
- Establish operational routines, dashboards, forecasts, and executive reporting for TaaS initiatives.
- Design scalable deployment models to enable future growth in industrial compute.

